Data imbalance and large data probability distribution discrepancies are major factors that reduce the accuracy of remaining useful life (RUL) prediction of high-reliability rotating machinery. In feature extraction, most deep transfer learning models consider the overall features but rarely attend to the local target features that are useful for RUL prediction; insufficient attention paid to local features reduces the accuracy and reliability of prediction. By considering the contribution of input data to the modeling output, a deep learning model that incorporates the attention mechanism in feature selection and extraction is proposed in our work; an unsupervised clustering method for classification of rotating machinery performance state evolution is put forward, and a similarity function is used to calculate the expected attention of input data to build an input data extraction attention module; the module is then fused with a gated recurrent unit (GRU), a variant of a recurrent neural network, to construct an attention-GRU model that combines prediction calculation and weight calculation for RUL prediction. Tests on public datasets show that the attention-GRU model outperforms traditional GRU and LSTM in RUL prediction, achieves less prediction error, and improves the performance and stability of the model.

Given the challenges of rotating equipment fault diagnosis under variable working conditions, including the unbalanced transmission of information during feature extraction, difficulty in capturing both global and local features, and limited generalization across different working conditions, a Chebyshev graph convolutional neural network (ChebyNet) method is proposed to address these issues. First, a symmetry processing mechanism is incorporated into the framework of the ChebyNet to balance the transfer of information between nodes in the graph to ensure the fair and efficient integration of information. Secondly, the wide-area feature extraction capabilities of the ChebyNet and the adaptive nodes of the graph attention network (GAT) are integrated to achieve the comprehensive mining of fault characteristics and accurate characterization of complex interactive relationships. Finally, the node reconstruction task of self-supervised learning and collaborative node classification tasks are used to enhance the model’s ability to capture complex changes in variable working conditions data, significantly improving the generalizability of working conditions. In comparative and cross-validation experiments, the proposed method achieved an average diagnostic accuracy of 99.72%, representing an improvement of up to 17.96% compared to other graph neural network (GNN) models. It significantly enhances the accuracy, stability, and generalization of fault diagnosis. Ablation experiments further validate the effectiveness of the proposed method in improving fault diagnosis performance under variable working conditions.

The rotating parts of large and complex equipment are key components that ensure the normal operation of the equipment. Accurate fault diagnosis is crucial for the safe operation of these systems. To simultaneously extract both local and global valuable fault feature information from key components of complex equipment, this study proposes a fault diagnosis network model, named MultiDilatedFormer, which is based on the fusion of transformer and multi-head dilated convolution. The newly designed multi-head dilated convolution module is sequentially integrated into the transformer-encoder architecture, constructing a feature extraction module where the complementary advantages of both components enhance overall performance. Firstly, the sample is expanded into a two-dimensional feature map and then input into the newly designed feature extraction module. Finally, the diagnostic output is performed by the designed patch feature fusion module and classifier module. Additionally, interpretability research is conducted on the proposed model, aiming to understand the decision-making mechanism of the model through visual analysis of the entire decision process. The experimental results on three different datasets indicate that the proposed model achieved high accuracy in fault diagnosis with relatively short data windows. The highest accuracy reached 97.95%, which was up to 10.97% higher than other models. Furthermore, the feasibility of the model is also verified in the actual dataset of the rotating parts of the injection molding machine. The excellent performance of the model on different datasets demonstrates its effectiveness in extracting comprehensive fault feature information and also proves its great potential in practical industrial applications.

Aiming at the sound source localization of mechanical faults in a strong reverberation scenario with multiple sound sources, this paper investigates a mechanical fault source localization method using the U‐net deep convolutional neural network. The method utilizes the SRP‐PHAT algorithm to calculate the response power spectra of the collected multichannel fault signals. Through the utilization of the U‐net neural network, the response power spectra containing spurious peaks are transformed into “clean” estimated source distribution maps. By employing interpolation search, the estimated source distribution maps are processed to obtain location estimations for multiple fault sources. To validate the effectiveness of the proposed method, this paper constructs an experimental dataset using mechanical fault data from electromechanical equipment relays and conducts sound source localization experiments. The experimental results show that the U‐net network under 0.2 s/0.5 s/0.7 s reverberation time can effectively eliminate spurious peak interference in the response power spectrum. As the signal‐to‐noise ratio decreases, it can still distinguish the sound sources with a distance of 0.2 m. In the context of multifault source localization, the method is capable of simultaneously locating the positions of four fault sources, with an average localization error of less than 0.02 m. The method in this paper effectively eliminates spurious peaks in the response power spectra under conditions of multisource strong reverberation. It accurately locates multiple mechanical fault sources, thereby significantly enhancing the efficiency of mechanical fault detection.

Reliable structural health monitoring and risk-informed maintenance of civil infrastructure systems rely on accurate life-cycle assessment of degradation and remaining useful life (RUL) of critical rotating machinery components. In such systems, rolling bearings are among the most failure-sensitive and maintenance-critical units. Rotating machinery, which is widely used in civil engineering systems for functions such as pumping, ventilation, and energy conversion, is particularly susceptible to performance degradation and failure. Owing to material fatigue, local damage evolution, and environmental disturbances, key performance indicators often exhibit nonlinearity, individual variability, and time-varying fluctuations. As a result, the degradation process of rotating machinery becomes highly uncertain, posing critical challenges for reliable RUL prediction. To address this issue, this paper proposes a RUL prediction method that integrates tensor decomposition with a Bayesian neural network (BNN). First, Bayesian tensor decomposition is applied to perform dimensionality reduction and feature extraction on high-dimensional monitoring data, thereby identifying key latent degradation features that form the basis for accurate prediction. On this basis, a probabilistic modeling framework is established, in which Bayesian tensor decomposition characterizes the variability of the extracted features and provides data-driven uncertainty estimates as prior information for the BNN. The extracted features are then fed into the BNN to achieve probabilistic RUL prediction while explicitly accounting for data uncertainty. Experimental results showed that the proposed approach mitigates interference from varying operating conditions, captures degradation trends, and improves prediction accuracy and robustness with reliable uncertainty quantification. These advantages highlight the potential of the proposed framework for supporting condition-based maintenance, risk-informed decision-making, and life-cycle reliability management of civil infrastructure systems in practical applications.